Table of Contents

Training Flow

This guide is for the trainers who deliver DataMiner Empower. It describes the full training flow and gives one guide per session with the flow, what is given, and the do's and don'ts.

Section For
Trainer Guide (this section) How to run each session
Participant Enablement What participants need and what we provide
Backend Packages The Catalog packages participants deploy for the hands-on sessions
Content Creation Everything that must be produced before the training

The Flow

flowchart LR
    K[1. Keynote<br/>Vision + Vanguard] --> D[2. DevOps Agents<br/>demo]
    D --> AB[3. App Builder<br/>hands-on]
    AB --> SA[4. Specialized Agents<br/>hands-on]
    SA --> A2A[5. Agent-to-Agent<br/>demo]
# Session Format Participants create Key takeaway Duration
1 Keynote: Vision + Vanguard Theoretical – Operations are heading towards AI-native, value-stream-driven work to be defined
2 DataMiner DevOps Agents Demonstration – Intent becomes implementation – the value is accelerated solution creation to be defined
3 App Builder Workshop Hands-on An application Solve an operational problem, not an infrastructure problem to be defined
4 Specialized Agents Workshop Hands-on An operational agent The most valuable agents understand how your organization operates to be defined
5 Agent-to-Agent Collaboration Demonstration – Realistic collaboration with controlled actions and human approval to be defined

The sessions build on each other: participants keep the same vertical from the App Builder Workshop into the Specialized Agents Workshop, and the Agent-to-Agent demo uses the same kind of operational context.


Core Philosophy

The training does not teach App Builder, DevOps Agents, DOM, ticketing, or AI agents as individual features. It creates a mindset shift.

Participants should stop thinking... ...and start thinking
"What application should I build?" "What operational outcome do I want to achieve?"

Every session reinforces the same cycle:

flowchart LR
    U[Understand] --> A[Act] --> C[Create]
    C -.->|Measure value & re-prioritize| U
Step Meaning Question to ask participants
Understand The operational problem, constraints, and desired future state What hurts today, and what does "good" look like?
Act Prioritize the highest-value opportunities Which opportunity returns the most value right now?
Create Only what is needed to prove value and achieve measurable outcomes What is the smallest thing that proves it works?

Key Messages to Reinforce in Every Session

1. Start with the outcome

Begin every session with these questions – never with technology:

  • What problem are we solving?
  • What does success look like?
  • What would the future state look like?
  • What business value would be created?
  • How do we reduce operational effort, cost, or risk?
  • How do we accelerate time-to-value?

2. Work backwards from the future state

  1. Define an ambitious future state
  2. Work backwards
  3. Identify the smallest valuable step
  4. Deliver something useful quickly
  5. Measure value
  6. Re-prioritize continuously

3. Multiple value streams

Participants may start one value stream and discover another with a much higher return. That is expected.

Value Stream Next Increment Value
A – Incident Resolution Current next increment Save 5 minutes per ticket
B – Operational Briefing Potential first increment Save 45 minutes per operator per day

→ Value Stream B should be prioritized.

We don't continue because we started. We continue because it is still the highest-value investment.

4. Operational data is the competitive advantage

The value is not AI. The value is operational telemetry, historical data, operational processes, company standards, policies, procedures, and institutional knowledge.

AI amplifies operational knowledge. Without a foundation of trusted operational data and knowledge, AI has limited value.


General Do's and Don'ts

Do Don't
Open every session with the outcome, not the tool Start with a feature tour
Use operational language (decisions, incidents, customers, risk) Use product jargon as the main message
Ask "Which decision does this help?" Reward complexity or visual polish
Celebrate small, working, valuable results Aim for completeness
Connect each session to the previous one Treat sessions as separate product trainings
Keep AI grounded in data and policies Present AI as magic or fully autonomous

How Each Session Guide Is Structured

Part Content
At a glance Format, what participants create, key message
Objective What the session must achieve
What is given Materials, environment, and data available in the session
Before the session Trainer preparation
Session flow Step-by-step: what the trainer does and what participants do
Do's and don'ts Session-specific guidance
Success criteria How you know the session worked
Transition How to hand over to the next session

Final Training Message

Participants should not leave saying "I learned App Builder" or "I learned how to create an agent."

They should leave saying:

I learned how to identify operational value, prioritize the most impactful opportunity, and rapidly create a working solution that improves operations using DataMiner's operational data, workflows, and organizational knowledge.